the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Apr 2026
Orographic impacts of Réunion Island and Madagascar on heavy rainfall during Tropical Cyclone Batsirai (2022)
Abstract. During the passage of tropical cyclone (TC) Batsirai (2022), the mountainous regions of Réunion Island and the eastern part of Madagascar experienced devastating floods. TC Batsirai passed north of Réunion Island (minimum distance of ~210 km) before making landfall along the eastern coast of Madagascar. Using the Meso-NH model at 2 km resolution, we conducted a control simulation with realistic topography (CTL) and three flattened-terrain experiments (FLTall, FLTmdg, and FLTreu) to assess orographic effects. Results show that the terrain of Réunion and Madagascar enhanced localized rainfall by 94 % (2,800 mm) and 82 % (830 mm), respectively. Over Réunion Island, orographically uplifted low-level moisture sustained 60 h of heavy rainfall (10 mm h‒1), with a peak of 96 mm h‒1, although the cyclone’s intensity and track were minimally affected. In Madagascar, terrain not only intensified precipitation (19 h of heavy rainfall with a peak intensity of 92.8 mm h‒1) but also altered the TC trajectory, delaying landfall by about 12 h and shifting it approximately 30 km south. These findings highlight the critical role of complex orography in shaping localized rainfall and TC behaviour in the southwest Indian Ocean.
- Preprint
(4217 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 30 Jun 2026)
- RC1: 'Comment on egusphere-2026-1893', Anonymous Referee #1, 08 Jun 2026 reply
-
RC2: 'Comment on egusphere-2026-1893', Anonymous Referee #2, 15 Jun 2026
reply
This paper investigates the orographic effects of a small island (Réunion Island) and a medium-sized island (Madagascar) on tropical cyclone (TC) track, intensity, and precipitation. The authors conducted a control experiment together with three sensitivity experiments with modified orography. The results show that the orography of Réunion Island has little impact on the TC track, while it plays an important role in controlling heavy precipitation over the island. In contrast, the orography of Madagascar slightly modifies the TC track and contributes significantly to heavy rainfall on the windward side of the island. The study clearly demonstrates the influence of island topography on TC behavior and associated precipitation. These findings are important for improving our understanding of orographic enhancement of TC-induced rainfall.
Overall, the manuscript is logically organized, and the scientific results are clearly presented. The reviewer finds the study valuable and believes that the manuscript has the potential to be accepted for publication. However, several issues should be addressed before publication. The reviewer kindly requests that the authors consider the following comments in revising the manuscript.
Major issue:
Lines 318-333: In this study, the authors compare the CTL experiment with the FLTall experiment and conclude that the peak precipitation and the duration of heavy rainfall are significantly reduced in FLTall. However, the reviewer believes that the comparison should be made between CTL and FLTreu rather than between CTL and FLTall. As shown in Figure 6, the storm tracks in CTL and FLTreu are very similar, whereas those in CTL and FLTall differ by approximately 40 km. According to Figures 9 and 10, such a difference appears to be important because it changes the relative position of Réunion Island with respect to the storm structure. In CTL, Réunion Island is located beneath the outer rainband, whereas in FLTall it is situated between the outer and inner rainbands. This suggests that the precipitation weakened more rapidly in FLTall, resulting in a shorter duration of heavy rainfall. Since CTL and FLTreu exhibit similar tracks and similar locations of the outer rainband, it is expected that the persistence of precipitation in FLTreu would be comparable to that in CTL. Therefore, the reviewer kindly recommends that FLTreu, rather than FLTall, be used for the comparison to isolate the effect of the orography of Réunion Island.
Specific points:
- Lines12-13: The statement that “the terrain of Réunion and Madagascar enhanced localized rainfall by 94% (2800 mm) and 82% (830 mm), respectively” is unclear. Does this imply that the increase in rainfall from the FLTxxx experiments to the CTL experiment corresponds to 94% and 82% of the rainfall amounts in the FLTxxx experiments? If so, this interpretation does not appear to be consistent with the presented results. The reviewer kindly requests that the authors clarify the basis of these percentages and explain more explicitly what is meant by these values.
- Line 169: What is “a monotonic Piecewise Parabolic Method”? a brief explanation is necessary for this.
- Line 199: The expression “the lee side of Réunion” is ambiguous. A more specific description should be provided to clearly indicate the region being referred to.
- Lines 267-270: The authors attribute the discrepancy between the observed and simulated precipitation amounts to an overestimation in the CTL simulation. However, another possible explanation should also be considered. The rain gauges located near the summit region of Réunion Island may have suffered from reduced catch efficiency under strong wind conditions, which could have resulted in an underestimation of the observed precipitation.
- Lines 331-333: This issue may be related to the major comment above. Why does the CTL experiment produce prolonged precipitation after 4 February? The authors should provide a physical explanation for the persistence of rainfall during this period.
- Lines 426-427: This issue may also be related to the major comment above. The absence of convective regions in FLTall may be attributed to the different relative position of Réunion Island with respect to the outer spiral rainband compared with CTL.
- Line 454: The “minimum difference” may be “maximum difference”.
- Lines 609-610: If the authors claim that equatorial Rossby waves play a dominant role in TC genesis, they should provide supporting evidence for this statement.
Citation: https://doi.org/10.5194/egusphere-2026-1893-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 311 | 122 | 22 | 455 | 15 | 16 |
- HTML: 311
- PDF: 122
- XML: 22
- Total: 455
- BibTeX: 15
- EndNote: 16
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Comment on “Orographic impacts of Réunion Island and Madagascar on heavy rainfall during Tropical Cyclone Batsirai (2022)” by Lee et al., 2026
The manuscript evaluatesthe impact of the orography on heavy precipitation, track and intensity of a typhoon in the Southwest Indian Ocean. The authorsuse the Meso-NH model in 2km resolution to conduct a control simulation with realistic topography, and three other runs with modified representation of the orography. The typhoon selected for the study is typhoon Batsirai that affected both Madagascar and Réunion island in 2022. Simulations outputs were compared to observational datasets including two C-band Doppler radars and rain gauges. Results show that orography enhanced precipitation over both islands and, in the case of Madagascar, also affected typhoon track and intensity.
The manuscript is well designed, clearly states the objectives and methodology, and the results are properly presented and discussed. Besides, it is important to the community due to its contribution to highlight the impacts of complex orography on typhoon track, intensity and precipitation over an area that has not received as much attention as Eastern Asia or the North Pacific. I recommend the paper for publication, although some comments should be addressed first.
General comments:
1. The authors mention that all simulations were initialized at 00 UTC on 1 February but did not mention the spin-up used to conduct these simulations. Later on, a 6 hour spin-up is mentioned in line 215 but these values are still used in the evaluation of the model. The authors should clarify which spin-up period was used and what the implications are of using it in the evaluation of the model. Is 6 hours of spin-up enough?
2. Regarding the sensitivity experiments, all terrain elevations were reduced to 10 m, but this decision is not justify in the text. Why 10 meters and not a different height?
3. Model validation is done using radar and rain gauges only for Réunion Island. Why not also for Madagascar? Besides, model validation is done only qualitatively and no evaluation metrics are shown.
4. The authors discuss the impact of orography on precipitation, track and intensity attributing all changes to the modified orography. However, changes in orography affect many relevant processes in the atmosphere such as radiation, convection, gravity waves or low-level thermally-induced circulation. For example, what are the impacts of orography changes on the latent heat flux? How is this linked with the simulated precipitation? The authors should discuss this in the manuscript.
Minor corrections
Line 24: Hauro et al., 1969 is not included in the references, or at least, not with year 1969. Please revise.
Line 64: “… showed that the presence of the island showed…” reformulate to avoid repetition of “showed”.
Line 94: it is the first time that IBTrACS is mentioned in the text. For better clarity, I would suggest to define its meaning briefly in the caption or to move the figure after the first time IBTrACS is mentioned in the text.
Line 128: Add in the caption what the black contour represents.
Line 145: “….Three-days accumulated from…”. Add rainfall before “from”.
Line 147: please, ad a link to SAR data either directly here or in the reference section.
Line 152: are the SAR-derived values used in this study corrected taking into account the bias mentioned in this line?
Line 153: why only values at 01:45 UTC on 4 February? I guess it is the only observation that fits into the dates and domain used for the study, but it should be explain just for the shake of clarity.
Line 162: is 1st of February used as spin up period? Taking into account figures shown in the rest o the manuscript, it seems that the authors used also this date in their evaluation and do not discard it. Could you please clarify this?
Line 172: there is no mention about the frequency of model outputs. Is it hourly. 6-hourly? Please specify.
Line 211: what does the green-dashed circle represent? There is no explanation in the caption.
Line 223: landfall time is mentioned here, but there is no label in Figure 3. Could you add it in the same fashion as for the other relevant dates shown there?
Line 252: why is the comparison done using different time for the model output than for the observation? And also why re wind profiles evaluated at 20 m?
Lines 271-274: I suggest to delete these lines as it has been already mentioned in lines 262-268 and it seems repetitive.
Figure 6: please include in the caption an explanation of what the dashed line represents, as well as regions I, ii and iii (currently there is no mention to it in the caption) and also what figures a) and b) show. It is also a bit confusing to the reader that Figure 6 shows the two plots together but Figure 6b is not mentioned until page 20 when the authors focus their attention on Madagascar. Perhaps, it should be mentioned in the caption that Figure 6b will be explained later in the Madagascar section.
Line 311: “...days; blueish areas...”. Blueish is not a formal word for a manuscript. Please replace it with blue shades for example.
Line 319: “(not shown)”. I am aware that the manuscript already has many figures, but it would be beneficial to see an example of the others changes in the orography, perhaps as supplemental material.
Figure 7: I suggest to show here just the differences and move the values of CTL to the supplement.
Figure 8: replace “Reunion” with “Réunion” for consistency.
Line 345: seems to be a space between “regio” and “n”.
Line 348: in order to make it clear to the reader, please specify that the grey-shaded areas are just for the control simulation.
Figure 9 and Figure 10: I suggest to combine Figure 9 and 10 into one single plot showing differences in the reflectivity between both simulations and then keep the plots showing low-level moisture flux as they are. Plots with the simulated reflectivity could be included in the supplement. Besides, the line with the cross section is difficult to see in its current color and the dashed lines in figure 9 are not mentioned in the caption. Please, add a brief explanation for clarity.
Line 373: include a brief description of the Froude number is and how it is calculated.
Line 377: change “In contract” for “In contrast”.
Figure 11: change the color of the topography to brown, for example, to avoid having similar color as those included in the color bar for potential temperature.
Line 483: “redish region...”. Same comment as in line 311.
Figure 12: the black dashed line is difficult to see now. Please consider changing color of thickness. For Figure 12 and 13, I suggest to do the same as I suggested for Figures 9 and 10. but here mentioning that the times taking for the comparison between CTL and FLT_all are different despite showing the same location.
Figure 14: Same comment for the color of the topography as in Figure 11. The dashed line is also difficult to see. Please change its color or thickness.
Line 611: it would be interesting to see how the other modes affect the genesis of the typhoon. Perhaps in the supplement.
Line 660: the DOI here does not belong to this citation.